Zoologica Scripta

Complete mitochondrial genomes throw light on budding speciation in three species (, Geometridae)

RUI CHENG,DAYONG XUE,ANTHONY GALSWORTHY &HONGXIANG HAN

Submitted: 22 December 2015 Cheng, R., Xue, D., Galsworthy, A. & Han, H. (2017). Complete mitochondrial genomes Accepted: 13 March 2016 throw light on budding speciation in three Biston species (Lepidoptera, Geometridae). — doi:10.1111/zsc.12184 Zoologica Scripta, 46,73–84. , Biston thibetaria and Biston perclara are very closely related species in the genus Biston, judged on the basis of both morphological and molecular evidence. The distri- bution area and altitudes, and host plants of these three species also show both consistency and differences. However, the exact relationship between the three species is unclear. In this study, we used the ‘distance-based method’, the ‘tree-based method’ and Bayesian phyloge- netics and phylogeography to elucidate the relationship between the three species. Phyloge- netic trees based on mitogenomes, COI+CYTB+16S and three nuDNA genes were constructed. The results of the phylogenetic trees revealed that B. thibetaria and B. perclara were derived from B. panterinaria and render the latter paraphyletic. The budding process of speciation is therefore presumed to be the main factor causing a phylogenetic relationship of this pattern. A host shift from broad-leaved plants living at low altitudes to gymnosperms living at relatively high altitudes provides evidence of budding speciation in these three spe- cies. The divergence time suggests that the budding speciation occurred at approximately 1.38 Ma, which is consistent with the Kunlun-Yellow River Movement. Tectonic move- ments occurring around the Qinghai–Tibet Plateau may be the driver of the budding speci- ation. Corresponding author: Hongxiang Han, Key Laboratory of Zoological Systematics and Evolution, Institute of Zoology, Chinese Academy of Sciences, No. 1 Beichen West Road, Chaoyang District, Beijing 100101, China. E-mail: [email protected] Rui Cheng, Key Laboratory of Zoological Systematics and Evolution, Institute of Zoology, Chinese Academy of Sciences, No. 1 Beichen West Road, Chaoyang District, Beijing 100101, China and University of Chinese Academy of Sciences, 19 Yuquan Road, Shijingshan District, Beijing 100049, China. E-mail: [email protected] Dayong Xue, Key Laboratory of Zoological Systematics and Evolution, Institute of Zoology, Chinese Academy of Sciences, No. 1 Beichen West Road, Chaoyang District, Beijing 100101, China. E-mail: [email protected] Anthony Galsworthy, The Natural History Museum, Cromwell Road, London SW7 5BD, UK. E-mail: [email protected] Hongxiang Han, Key Laboratory of Zoological Systematics and Evolution, Institute of Zoology, Chinese Academy of Sciences, No. 1 Beichen West Road, Chaoyang District, Beijing 100101, China. E-mail: [email protected]

Introduction 1970; White 1978; Futuyma & Mayer 1980; Coyne & Orr The origin of species is a fundamental issue in evolution 2004; Nosil 2012). Budding speciation is one specific type biology. Most biologists agree that speciation should be of peripatric speciation, which, as a concept, was first pro- seen as the evolution of reproductive isolation between posed by Mayr (1954): he envisaged an initially small colo- populations (Coyne 1994; Coyne & Orr 1998, 2004; Noor nizing population becoming reproductively isolated from 2002). Many kinds of speciation have been described, the wider-ranging parental species. However, the proposal including allopatric speciation, parapatric speciation, peri- of this type of speciation did not at first draw attention. patric speciation and sympatric speciation (Mayr 1942, Later, Wiley & Mayden (1985), Harrison (1991, 1998),

ª 2016 Royal Swedish Academy of Sciences, 46, 1, January 2017, pp 73–84 73 Budding speciation in three Biston species  R. Cheng et al.

Frey (1993) and Rieseberg & Brouillet (1994) elaborated (nuDNA) loci, to test the paraphyly of B. panterinaria and the concept of peripheral isolates. In particular, Harrison its consistency with the hypothesis of budding speciation. If (1991) identified peripheral isolates based on molecular evi- this result is deemed authentic, these three species will dence, with species A embedding within paraphyletic spe- form a good case study for studying budding speciation in cies B. In 2003, Funk & Omland improved the concept of Lepidoptera. budding speciation and listed much evidence and interpre- tation, following which the concept of budding speciation Material and methods began to be widely accepted. Funk & Omland defined bud- Specimen sampling and DNA extraction ding speciation, or peripatric and peripheral isolates, as The sampling sites of B. panterinaria, B. thibetaria and always implying a geographically restricted and mono- B. perclara are shown in Fig. 1. Samples for DNA extrac- phyletic daughter species being embedded within a widely tion were preserved in 100% ethanol and stored at distributed and paraphyletic parental species. Subsequently, À20 °C. DNA was extracted using the DNeasy Tissue kit many examples of budding speciation were discovered, (Qiagen, Beijing, China), and vouchers were deposited at involving liverworts, kelps, flowers, , snails, frogs and the Museum of IZCAS (the Institute of Zoology, Chinese shrews (Morse & Farrell 2005; Vanderpoorten & Long Academy of Sciences, Beijing, China). Ten mitogenomes 2006; Chen et al. 2009, 2012; Tellier et al. 2009; Grossen- from four lineages of B. panterinaria (Cheng et al. 2015), bacher et al. 2014; Kruckenhauser et al. 2014). three mitogenomes from B. thibetaria and one mitogenome Biston panterinaria (Bremer & Grey, 1853), Biston thi- from B. perclara were obtained. The mitogenomes were betaria (Oberthur,€ 1886) and Biston perclara (Warren, 1899) amplified as described in Yang et al. (2013). Three mito- have since their original description been regarded as chondrial genes and three nuclear genes were obtained specifically valid members of the genus Biston of the family from total specimens from the above-mentioned three spe- Geometridae, Lepidoptera, and are distributed in east Asia, cies and from ten other Biston species, including the cyto- south China and Taiwan, respectively. Importantly, these chrome c oxidase subunit I (COI), cytochrome b (CYTB), three species are closely related species forming Group III lrRNA (16S), elongation factor 1a (EF-1a), wingless (wg) of the genus based on morphological characters (Jiang et al. and glyceraldehyde-3-phosphate dehydrogenase (GADPH), 2011). In addition, the distributional areas and altitudes, through polymerase chain reaction (PCR) amplification. and host plants of these three species show both consis- PCR conditions were as follows: denaturation at 94 °C for tency and differences. The host plants of B. panterinaria 2 min, followed by 40 cycles at 93 °C for 1 min, annealing are mainly broad-leaved arbours (Yu 2001; Zhang & Li temperature (51 °C for COI,47°C for CYTB,56°C for 2001; Cui 2004; Li et al. 2008) below 1500 m, while the EF-1a,58°C for wg and 55 °C for GAPDH) for 45 s, and host plants of B. thibetaria are mainly conifers belonging to 72 °C for 1 min, and a final 10 min at 72 °C. PCR condi- the gymnosperms (Chen et al. 1982; Zheng & Yang 1991) tions for 16S was as follows: denaturation at 94 °C for from 1000 to 3000 m. However, the precise relationship 3 min, followed by 15 cycles at 94 °C for 30 s, annealing between the three species is unclear. temperature at 55–60 °C for 30 s, 72 °C for 1 min, fol- During our work of phylogeographic analysis of B. pan- lowed by 25 cycles at 94 °C for 35 s, annealing tempera- terinaria, we found that the relationships between B. pan- ture at 50 °C for 30 s, 72 °C for 1 min, and a final 10 min terinaria, B. thibetaria and B. perclara were unusual and that at 72 °C. Multiple alignments were generated using B. panterinaria seemed paraphyletic. However, we were not CLUSTALX (Thompson et al. 1997). Sequences of all pri- sure whether this phenomenon was authentic, because the mers used in this study are listed in Appendix S1. sampling numbers were small. So, in this study, we used Sequences were deposited in GenBank; the accession num- more samples, more data sets and more methods to explore bers are provided in Appendix S2. the relationships of the three species in detail. Nowadays, complete mitochondrial genomes (mitogenomes) can pro- Calculating interspecific sequence divergence and duce more robust and stable phylogenetic reconstructions constructing the NJ tree than those which rely on only part of the mtDNA, and ‘Distance-based method’ and ‘tree-based method’ were have been increasingly applied to a variety of phylogenetic used to ascertain the relationship of the three Biston spe- and phylogeographic studies in invertebrates (Ingman et al. cies. All COI sequences of the three species were used to 2001; Ma et al. 2012; Cameron 2014; Liu et al. 2015; Yang calculate the interspecific sequence divergence and con- et al. 2015). So the phylogenetic relationships of B. panteri- struct the neighbour-join (NJ) tree by MEGA 5.05 (Tamura naria, B. thibetaria and B. perclara were reconstructed based et al. 2011) based on Kimura-2-parameter (K2P) distances. on three data sets, involving mitogenomes, three mitochon- NJ analysis (Saitou & Nei 1987) was used to examine drial DNA (mtDNA) loci and three nuclear DNA whether these three species were effective and authentic.

74 ª 2016 Royal Swedish Academy of Sciences, 46, 1, January 2017, pp 73–84 o C

R. Cheng et al.  Budding speciation in three Biston species

Fig. 1 Sampling sites of B. perclara, B. thibetaria and B. panterinaria. Colours represent different species.

Species tree estimation three species to obtain important characters. Morphological A species tree of the three species was constructed using taxonomic methods were used to examine external features Bayesian phylogenetics and phylogeography (BPP v.2.0; and genitalia preparations. The terminology of the genitalia Yang & Rannala 2010) with the full phased data set for the follows that of Pierce (1914, reprint 1976), Klots (1970) and three mitochondrial loci. The model assumes no admixture Nichols (1989). Photographs of the were taken with following speciation, which is an assumption motivated by digital cameras. Composite images of genitalia were gener- the biological species concept. The modes of rjMCMC ated using Auto-Montage 5.03.0061 (Synoptics, Ltd, Cam- species delimitation and NN1 over species/guide trees were bridge, Cambridgeshire, UK). The plates were compiled chosen. Running the rjMCMC analyses for 500 000 gener- using ADOBE PHOTOSHOP SOFTWARE (San Jose, CA, USA). ations (sampling interval of five) with a burn-in period of 10 000 produced consistent results across separate analyses Phylogenetic analysis of Biston initiated with different starting seeds. We ran each analysis To make sure of the close relationship of these three spe- ten times with different starting seeds to check conver- cies according to the molecular evidence, we constructed a gence by examining agreement between runs. Ensuring phylogenetic tree of Chinese Biston based on combined adequate rjMCMC mixing involves specifying a reversible mtDNA and nuDNA genes. One or two specimens of jump algorithm to achieve dimension matching between every species were chosen. Two methods of Bayesian infer- species delimitation models with different numbers of ence (BI) and maximum likelihood (ML) analysis were parameters, and we used algorithm 0 with the fine-tuning used. parameter ɛ = 2. Each species delimitation model was For the BI analysis, the best-fit model of nucleotide sub- assigned equal prior probability. We evaluated the influ- stitution was selected using JMODELTEST 0.1.1 (Guindon & ence of the priors for population sizes (h) and the root Gascuel 2003; Posada 2008) under the Bayesian informa- height of the guide tree (s) on the posterior probability of tion criterion (BIC) (Schwarz 1978). BI analysis was carried species delimitation models by considering different combi- out in MRBAYES 3.1.2 (Huelsenbeck & Ronquist 2001; nations of priors. Ronquist & Huelsenbeck 2003) under default priors, with each partition unlinked for parameter estimations. Four Comparing the characters of female genitalia Markov chains (three heated and one cold) were run, start- In earlier taxonomic work on moths, the male genitalia were ing from a random tree and proceeding for 1 000 000 often used as a main basis for classification, a method which Markov chain Monte Carlo (MCMC) generations, sam- was accepted by almost all lepidopterists. As a result, little pling the chains every 100 generations. Two concurrent information about the female genitalia can be found in the runs were conducted to verify the results. A plot of sam- literature. In our study, as the male genitalia of the three Bis- pled log-likelihood scores against generation time was used ton species showed few differences, so we compared the to determine the stationarity of the chains. The trees prior female genitalia of the main geographical populations of the to stationarity were discarded. For all runs, the first 2500

ª 2016 Royal Swedish Academy of Sciences, 46, 1, January 2017, pp 73–84 75 Budding speciation in three Biston species  R. Cheng et al. trees were discarded as burn-in samples. The remaining put. We used TreeAnnotator in the BEAST package to trees were used to compute a majority-rule consensus tree summarize tree data with ‘mean height’ and discarded the with posterior probabilities (PP) (for the overcredibility of first 25% of trees as the ‘burn-in’ period, which ended well Bayesian phylogenetics, see Suzuki et al. 2002; Cummings after the stationarity of the chain likelihood values that had et al. 2003; Simmons et al. 2004). been established. The tree and divergence times are The maximum likelihood (ML) analysis was inferred in displayed in FIGTREE 1.3.1. RAXML v7.2.6 (Stamatakis 2006; Stamatakis et al. 2008) with the GTRGAMMAI model for each partition. All Results model parameters were estimated during the ML analysis. Interspecific sequence divergence and NJ tree of three Biston A rapid bootstrapping algorithm with a random seed value species of 12345 (command -f a -x 12345) was applied with 1000 The results of interspecific divergence of these three spe- replicates (Siddall 2010). cies are listed in Appendix S3. The divergences between B. panterinaria and B. thibetaria, B. panterinaria and Phylogenetic analysis of three Biston species B. perclara, B. perclara and B. thibetaria are 3.9–5.2%, 3.7– Phylogenetic analysis of B. panterinaria, B. thibetaria and 5.6% and 3.6–4.3%, respectively. The results of the NJ B. perclara was constructed based on three data sets (mi- analysis based on COI gene are summarized in Fig. 2. In togenomes, COI+CYTB+16S and each nuDNA gene). The the NJ tree, four lineages of B. panterinaria formed one mitogenome of (NC_027111) was used distinct cluster. B. perclara and B. thibetaria formed another as an outgroup. Two methods of BI and ML were used cluster. according to the above-mentioned instructions. The data of mitogenomes were divided into 15 partitions (13 PCGs Species tree and 2 rRNA genes) and 41 partitions (13 PCGs partitioned The results of BPP are shown in Fig. 3. The topology of by codon position and 2 rRNA genes partitioned by gene). the species tree is congruent with that of the NJ tree. Four Different models were used for different partitions, which lineages of B. panterinaria grouped into one clade, and were selected using the jModelTest. Poorly aligned nucleo- B. thibetaria and B. perclara grouped into one clade, respec- tide positions were omitted by the program GBLOCKS (Cas- tively. tresana 2000) with a more stringent selection criterion. The results were visualized using FIGTREE 1.3.1 (Rambaut The characters of female genitalia of three Biston species 2009). The adults and female genitalia of the three species are shown in Fig. 4. Some characters, such as ostium bursae Estimate of divergence time and signum, are very close among the three species. But The maximum clade credibility tree from divergence-time- they differ substantially in wing markings, and in the shape rooted phylogenetic analyses was estimated using BEAST and degree of sclerotization of the lamella postvaginalis. 1.8.0 (Drummond & Rambaut 2007) based on COI+- The lamella postvaginalis is triangular, weakly to moder- CYTB+16S data set. Although the use of a molecular clock ately sclerotized in B. panterinaria, whereas, it is tongue- as the only way for calibrating the divergence time of phy- like, and well sclerotized in B. thibetaria and B. perclara, logenetic trees is controversial, it does provide a method and posteriorly narrower in B. perclara. for estimating approximate time when no other calibration information, such as fossil or geological evidence, is avail- The close relationship of three Biston species able (Maekawa et al. 2001; Lopez-L opez et al. 2015). Thus, A total of thirteen species from three species groups the widely accepted mutation rates for mitochondrial (groups I, II and III) based on morphological evidence COI gene (0.0115–0.0177 per site per million years, Brower were used to reconstruct a phylogenetic analysis of Chinese 1994; Papadopoulou et al. 2010) were adopted; other genes Biston. Phylogenetic trees based on combined mtDNA and were scaled to the COI rate in BEAST. Finally, the already nuDNA genes showed that Biston is divided into two clades established divergence time of the northern and Yunnan– (Appendix S4), which is not consistent with the three- Tibet lineage of B. panterinaria (Cheng et al. 2015) could group system based on morphology. One clade included be used to calibrate the results. Each gene was assigned a Group II (B. falcata, B. brevipennata and B. quercii) and separate unlinked relaxed clock model in the analysis. some species of Group I (B. regalis, B. thoracicaria and Default priors were used. Chains were analysed for 200 B. betularia). The other clade included Group III million generations, with sampling every 2000 generations. (B. perclara, B. thibetaria and B. panterinaria) and some TRACER 1.5.0 was used to verify the posterior distribution species of Group I (B. bengaliaria, B. suppressaria, B. medio- and the effective sample sizes (ESSs) from the MCMC out- lata and B. contectaria). The Group III of B. perclara,

76 ª 2016 Royal Swedish Academy of Sciences, 46, 1, January 2017, pp 73–84 R. Cheng et al.  Budding speciation in three Biston species

Fig. 3 Species tree estimated by BPP. Posterior probabilities of clade support are shown at nodes above branches.

clade including B. perclara and B. thibetaria and another clade including the northern and Yunnan–Tibet lineages of B. panterinaria grouped together. The phylogenetic trees of each nuDNA gene (Appendix S6) showed an incomplete lineage sorting of the three Biston species, which did not form stable clusters.

Divergence-time estimation The result of divergence-time analysis is shown in Fig. 7. The most recent common ancestor (TMRCA) of B thi- betaria and B. perclara separated from lineage I (including northern and Yunnan–Tibet lineages) of B. panterinaria at approximately 1.38 Ma [(1.04–1.81 Ma, 95% highest poste- rior density (HPD)], which is within the era of the Kun- lun-Yellow River Movement. The divergence time between B thibetaria and B. perclara is approximately 1.12 Ma (0.82– 1.50 Ma, 95% HPD).

Discussion The phylogenetic relationship of the Chinese Biston Fig. 2 NJ tree based on Kimura-2-parameter model for COI The genus Biston in China includes 17 species and has been sequence from B. perclara, B. thibetaria and B. panterinaria. Nodal divided into three species groups based on morphological values of bootstrap support are shown. characters (Jiang et al. 2011). Group I comprises the ‘typi- cal’ species of Biston. Group II comprises B. brevipennata B. thibetaria and B. panterinaria still grouped together, Inoue, 1982, and the species which were placed in the sub- which confirmed the particularly close relationship of these genus Eubyjodonta of Biston by Wehrli (1941). Group III three species. comprises B. panterinaria, B. thibetaria and B. perclara, which were considered somewhat different from the typical Phylogenetic analysis of the three Biston species Biston species by Sato (1996). This division does not accord The ML tree and BI tree based on mitogenomes and with the phylogenetic analysis based on molecular evidence. COI+CYTB+16S are essentially identical in topology, with The result of phylogenetic analysis based on multiple loci the exception of the position of the southern and Yunnan– divides Biston in China into two clades. One clade includes SE lineages of B. panterinaria (Figs 5 and 6 and Group II and some species of Group I. The other clade Appendix S5). All phylogenetic trees suggested that B. pan- includes Group III and some species of Group I. Impor- terinaria is not monophyletic and that B. perclara and tantly, B. panterinaria, B. thibetaria and B. perclara grouped B. thibetaria are embedded within B. panterinaria. The together and consisted of one subclade of the second clade.

ª 2016 Royal Swedish Academy of Sciences, 46, 1, January 2017, pp 73–84 77 Budding speciation in three Biston species  R. Cheng et al.

Fig. 4 Adults, female genitalia and enlarged lamella postvaginalis. A, B, C and D are northern, Yunnan–Tibet, southern and Yunnan–SE lineages of B. panterinaria, respectively; E: B. thibetaria;F:B. perclara.

This phylogenetic analysis confirmed the close relationship Appendix S5). In our previous study of B. panterinaria, four of B. panterinaria, B. thibetaria and B. perclara, which is lineages were reciprocally monophyletic, and the southern congruent with the morphology. A close relationship of sis- and Yunnan–SE lineages and the northern and Yunnan– ter species is one foundation of studying budding specia- Tibet lineages grouped together, respectively (Cheng et al. tion within the three species. 2015): which is different with the phylogenetic trees of three species. Among all phylogenetic trees of the three The relationship of the three Biston species Biston species, B. perclara, B. thibetaria and the northern In our study, B. panterinaria, B. thibetaria and B. perclara and Yunnan–Tibet lineages of B. panterinaria grouped into are confirmed as closely related species or sister species in one stable clade. It seems that B. panterinaria is not a the genus Biston based on morphological and molecular monophyletic species. evidence. They share many features in the genitalia, such Based on our phylogenetic results, we consider that the as the bifurcated uncus, short and apically rounded most reasonable conclusion is that B. perclara + B. thibetaria gnathos, and very short ductus bursae (Sato 1996; Jiang are derived from within B. panterinaria, rendering the latter et al. 2011). They cannot be distinguished by the male gen- paraphyletic. Based on the concept of budding speciation, italia, but can be distinguished by wing pattern and vena- we think that the common ancestor of B. perclara and tion, and especially by the shape and degree of B. thibetaria is the result of budding speciation from their sclerotization of the lamella postvaginalis in the female parent species (B. panterinaria) (Mayr 1954; Funk & genitalia. In addition, the NJ tree and species tree also sup- Omland 2003; Morse & Farrell 2005; Vanderpoorten & ported the existence of three good species, on the basis of Long 2006; Toussaint et al. 2013; Grossenbacher et al. four lineages of B. panterinaria, B. thibetaria and B. perclara 2014). The following evidence is relevant: (i) these three grouped as one clade respectively. species are most closely related species or sister species, Phylogenetic trees based on the mitogenomes and three which is one important basis of budding speciation mtDNA genes (COI+CYTB+16S) show a similar topology, (Grossenbacher et al. 2014); (ii) B. perclara + B. thibetaria with the exception of the relationship of the southern and are embedded in B. panterinaria in the phylogenetic trees, Yunnan–SE lineages of B. panterinaria (Figs 5 and 6 and which means B. panterinaria is a widely distributed and

78 ª 2016 Royal Swedish Academy of Sciences, 46, 1, January 2017, pp 73–84 R. Cheng et al.  Budding speciation in three Biston species

Fig. 6 Phylogenetic tree of B. perclara, B. thibetaria and Fig. 5 Phylogenetic tree of B. perclara, B. thibetaria and B. panterinaria based on COI+CYTB+16S data set. Values on the B. panterinaria based on mitogenomes. Values on the left of nodes left of nodes indicate bootstrap supports of ML for major clades. indicate bootstrap supports and posterior probabilities of ML/BI for major clades. interrupt gene flow between the various regions of distri- bution, whereas there are both altitudinal and food plant paraphyletic parental species and B. perclara + B. thibetaria characteristics which provide a convincing isolation mecha- are geographically restricted monophyletic daughter spe- nism between B. panterinaria and the other two species, cies; (iii) the distribution of B. perclara + B. thibetaria over- and in addition, there is geographical separation in the case laps with the distribution of B. panterinaria, which is of B. perclara. The existence of morphological differences congruent with the concept of peripatric speciation; (iv) the in the female genitalia between B. panterinaria (from right result of nuDNA analysis and male genitalia, and relatively across its range) on the one hand, and B. thibetaria and low genetic distance showed the effect of incomplete lin- B. perclara on the other point is in the same direction. eage sorting; (v) different host plants and different altitudes Moreover, the smaller genetic distances are exactly what provided little chance to do interspecific introgression would be expected where lineage sorting is incomplete. For (Funk & Omland 2003; McKay & Zink 2010). these reasons, it seems much more likely that B. panterinaria However, an alternative conclusion from the molecular represents a comparatively stable entity, albeit with genetic data might theoretically be that the four lineages of B. pan- variations, across its range, while the other two species are terinaria, B. thibetaria and B. perclara actually comprise isolated and diverging, but derived from within the B. pan- some four or five monophyletic taxa within the B. panteri- terinaria complex on the ‘budding speciation’ principle. naria complex, whether those taxa are considered to be valid at the specific or subspecific level, with the B. panteri- The process of budding speciation naria populations at either end of the distribution evolving In this study, the divergence time of the common ancestor into separate taxa. According to Hausmann et al. (2011 and of B. perclara + B. thibetaria and B. panterinaria is approxi- 2013) the average genetic distances between geometrid spe- mately 1.38 Ma (1.04–1.81 Ma, 95% HPD), which is con- cies are considerably higher, often between eight and 10%: sistent with the Kunlun-Yellow River Movement induced in this case, the differences are clearly smaller. This con- by the uplift of the Qinghai–Tibet Plateau (Cui et al. 1997, clusion would require the B. panterinaria clades on either 1998). So it is likely that the time of budding speciation side of the B. thibetaria +B. perclara clade to be treated as was during the period of the Kunlun-Yellow River Move- separate taxa. But we do not find this conclusion persuasive ment. In the phylogenetic tree, the common ancestor of for a number of reasons: the distribution of B. panterinaria B. perclara + B. thibetaria and the Yunnan–Tibet + north- feeding on broad-leaved foliage is fairly continuous across ern lineages of B. panterinaria clustered as one clade (Figs 5 its range, and there is no obvious mechanism which would and 6). The distribution of the Yunnan–Tibet and

ª 2016 Royal Swedish Academy of Sciences, 46, 1, January 2017, pp 73–84 79 Budding speciation in three Biston species  R. Cheng et al.

Fig. 7 Divergence-time-rooted phyloge netic analysis of B. perclara, B. thibetaria and northern and Yunnan–Tibet lineages of B. panterinaria based on COI+CYTB+ 16S data set. Posterior probabilities of clade support are shown at nodes above branches. Estimates of divergence time with 95% confidence intervals are shown at nodes as purple bars and numbers below branches. northern lineages of B. panterinaria is on the east slope tectonic movements induced by the uplift of the Qinghai– of the Gaoligong Mountains, south-eastern Tibet and Tibet Plateau were the main driver of this speciation pro- north China, and the distribution of B. perclara + B. thi- cess. The host shift from angiosperm to gymnosperm is betaria is the east slope of the Gaoligong Mountains, also found in other insects (Sequeira et al. 2000; Farrell south China and Taiwan Island. The common distribu- et al. 2001). tion is the east slope of the Gaoligong Mountains, which Budding speciation represents a situation where lineage therefore seems to be the most likely region of budding sorting is incomplete (Funk & Omland 2003; Vander- speciation. In addition, the population of Weixi County poorten & Long 2006; Enroth et al. 2009; Kruckenhauser (the possible westernmost distribution region of et al. 2014), which implies that the paraphyletic situation B. perclara + B. thibetaria) is located at the root of the will disappear and the genetic distance of the three species phylogenetic tree. So, after budding speciation, the com- will increase, as soon as the lineage is completely sorted. mon ancestor of B. perclara and B. thibetaria may have dispersed from west to east and eventually colonized Driver of budding speciation Taiwan Island. Although this form of speciation is still controversial, mul- In budding speciation, the newly formed small-range tiple reported examples, involving insects, snails, frogs and species should occupy a distinct realized niche when com- shrews (Morse & Farrell 2005; Chen et al. 2009, 2012; pared to their large-range sisters (Grossenbacher et al. Kruckenhauser et al. 2014), have improved the credibility 2014): it has been widely recognized that the majority of of the concept. Basing on the study of Mimulus (mon- speciation events in phytophagous insects were accompa- keyflowers), budding speciation seems to be very common nied by shifts in host plants (Ehrlich & Raven 1964; Mayr in certain taxonomic groups and in certain regions (Kruck- 1982; Mitter et al. 1988; Thompson 1994; Funk et al. 1995; enhauser et al. 2014). In these studies, one important char- Farrell 1998; Morse & Farrell 2005). Considering the big acter of budding speciation is that the newly emerging differences in diet between B. perclara, B. thibetaria and B. species, compared to parental species, always occupies dis- panterinaria, we think the likely sequence is that one popu- tinct and different niches, involving host plant divergence lation of B. panterinaria from the east slope of the Gaoli- (Morse & Farrell 2005), and adaptation to different habi- gong Mountains, previously feeding on broad-leaved plants tats and soils (Anacker & Strauss 2014; Kruckenhauser at low altitudes, adapted to feeding on gymnosperms at et al. 2014), colonizing other regions by long-distance dis- higher altitudes, following which reproductive isolation persal (Wiley & Mayden 1985; Vanderpoorten & Long occurred, resulting in the appearance of the distinct com- 2006). Budding speciation is driven by ecological opportu- mon ancestor of B. perclara and B. thibetaria during the nities offered by the emergence of largely unoccupied habi- Kunlun-Yellow River Movement. If this is so, it seems that tats with few competitors (Hughes & Eastwood 2006;

80 ª 2016 Royal Swedish Academy of Sciences, 46, 1, January 2017, pp 73–84 R. Cheng et al.  Budding speciation in three Biston species

Hines 2008). There are two important factors resulting in (Hewitt 2001; Qu et al. 2014; Lei et al. 2015). Such migra- new ecological opportunities: climatic fluctuations and tec- tion may well have provided a more continuous path, tonic movements. When climate changes (e.g. the presence enabling the common ancestor of B. perclara and B. thi- of an ice age), new environmental conditions (e.g. tempera- betaria, both gymnosperm feeders, to spread far enough ture, humidity), along with new host plants create ecologi- across southern China to reach Taiwan across the land cal opportunities (Morse & Farrell 2005; Vanderpoorten & bridge. After glaciation, the gymnosperms will have Long 2006; Toussaint et al. 2013). In addition, tectonic retreated again to higher altitudes, so that the dispersal movements create multiple new ecological opportunities route was cut off, and the became restricted to rare (Chen et al. 2009). The uplift of the Qinghai–Tibet plateau high altitude areas of southern China, especially the Heng- provided ecological opportunities in the unoccupied habi- duan Mountains, which is quite distant from Taiwan. The tats of higher altitudes (Cao et al. 1981). When organisms moth then evolved through the isolation mechanism into take full advantage of such opportunities, speciation and two distinct species at the two ends of its former range. In diversification can occur rapidly (Chen et al. 2009). The summary, the discontinuity of host plants may be the rea- conclusion from our study, with the uplift of the Qinghai– son for the current restricted distributions of B. perclara Tibet plateau, is that one population of B. panterinaria and B. thibetaria. from the east slope of Gaoligong Mountains occupied the feeding niche created by the proliferation of gymnosperms Acknowledgements at the high altitudes, which promoted the budding process We gratefully thank all collectors for their assistance in + of the common ancestor of B. thibetaria B. perclara. Our Lepidoptera collection. Special thanks to Dr. Shipher Wu, study is thus a good example of budding speciation driven Biodiversity Research Center, Academia Sinica, Taiwan, – by tectonic movements of the Qinghai Tibet plateau. for kindly providing the specimens of B. perclara. We thank Dr Malcolm Scoble, The Natural History Museum, Lon- The relationship of B. perclara and B. thibetaria don, UK, for his helpful comments. This work was sup- Based on morphology and molecular evidence, B. perclara ported by the National Science Foundation of China (No. and B. thibetaria are two sister species, which were isolated 31272288, 31372176 and 31572301), the National Science by the Taiwan Strait. This form of evolution is very com- Fund for Fostering Talents in Basic Research (No. NSFC- mon in almost all organisms, and the fauna of Taiwan J1210002) and a grant from the Key Laboratory of the Island has very close ties with that of the mainland of Zoological Systematics and Evolution of the Chinese Acad- China (Ying & Hsu 2002; Huang et al. 2004; Yen 2013). emy of Sciences (No. O529YX5105). Taiwan Island separated from the mainland at approxi- mately 5 Ma for the first time (Sibuet & Hsu 1997, 2004) References and then was rejoined through a land bridge during multi- Anacker, B. L. & Strauss, S. Y. (2014). The geography and ecology ple glacial periods and completely separated after the last of plant speciation: range overlap and niche divergence in sister glacial maximum (Voris 2000). The divergence time species. Proceedings of the Royal Society of London B: Biological between B. perclara and B. thibetaria is about 1.12 (0.82– Sciences, 281, 20132980. 1.50, 95% HPD) Ma (Fig. 7), which is close to the Xixia- Atkinson, P. R. (1980). On the biology, distribution and natural bangma glaciation (0.8–1.17 Ma, Shi 2002; Zheng et al. host-plants of Eldana saccharina Walker (Lepidoptera: Pyralidae). 2002). A reasonable conclusion would be that after the Xix- Journal of the Entomological Society of Southern Africa, 43, – iabangma glaciation, when the land bridge disappeared, and 171 194. Ballabeni, P., Conconi, D., Gateff, S. & Rahier, M. (2001). Spatial the Taiwan Strait again became a geographical barrier, gene proximity between two host plant species influences oviposition fl ow between these two populations ceased and the two and larval distribution in a leaf beetle. Oikos, 92, 225–234. populations evolved independently into separate species. Brower, A. V. Z. (1994). Rapid morphological radiation and con- The distribution range of phytophagous insects can be vergence among races of the butterfly Heliconius erato inferred expected to have a close relationship with their host plants from patterns of mitochondrial DNA evolution. Proceedings of the – (Atkinson 1980; Lance & Barbosa 1982; Jermy 1984; National Academy of Sciences, USA, 91, 6491 6495. Stoyenoff et al. 1994; Tikkanen et al. 1999; Ballabeni et al. Cameron, S. L. (2014). Insect mitochondrial genomics: implica- tions for evolution and phylogeny. Annual Review of Entomology, 2001; Lopez-Vaamonde et al. 2003). During glaciation, 59,95–117. conditions are created in which gymnosperms, which nor- Cao, W. X., Chen, Y. Y., Wu, Y. F. & Zhu, S. Q. (1981). Origin mally occupy high altitudes, can migrate to lower altitudes, and evolution of Schizothoracine Fishes in relation to the uphea- replacing broad-leaved vegetation which is less suited to val of the Qinghai-Xizang Plateau. In: Studies on the Period, the changed climatic conditions: this is a common strategy Amplitude and Type of the Uplift of the Qinghai-Xizang Plateau, fi for gymnosperms avoiding the influence of glaciation The Comprehensive Scienti c Expedition to the Qinghai-

ª 2016 Royal Swedish Academy of Sciences, 46, 1, January 2017, pp 73–84 81 Budding speciation in three Biston species  R. Cheng et al.

Xizang Plateau, Academia Sinica (pp. 118–130). Beijing: Science Funk, D. J. & Omland, K. E. (2003). Species-level paraphyly and Press. polyphyly, frequency, causes, and consequences, with insights Castresana, J. (2000). Selection of conserved blocks from multiple from mitochondrial DNA. Annual Reviews in Ecology and alignments for their use in phylogenetic analysis. Molecular Biol- Evolution, 34, 397–423. ogy and Evolution, 17, 540–552. Funk, D. J., Futuyma, D. J., Ortı, G. & Meyer, A. (1995). A his- Chen, E. H., Suo, Q. H. & Zeng, S. S. (1982). The survey on tory of host associations and evolutionary diversification for defoliator of Cupressus Duclouxiana in Chengjiang area. Yunnan Ophraella (Coleoptera, Chrysomelidae), new evidence from mito- Forestry Science and Technology, 2,89–92. chondrial DNA. Evolution, 49, 1008–1017. Chen, W., Bi, K. & Fu, J. (2009). Frequent mitochondrial gene Futuyma, D. J. & Mayer, G. C. (1980). Non-allopatric speciation introgression among high elevation Tibetan megophryid frogs in . Systematic Biology, 29, 254. revealed by conflicting gene genealogies. Molecular Ecology, 18, Grossenbacher, D. L., Veloz, S. D. & Sexton, J. P. (2014). Niche 2856–2876. and range size patterns suggest that speciation begins in small, Chen, S. D., Liu, S. Y., Liu, Y., He, K., Chen, W. C., Zhang, X. ecologically diverged populations in North American mon- Y., Fan, Z. X., Tu, F. Y., Jia, X. D. & Yue, B. S. (2012). Molec- keyflowers (Mimulus spp.). Evolution, 68, 1270–1280. ular phylogeny of asiatic short-tailed shrews, genus Blarinella Guindon, S. & Gascuel, O. (2003). A simple, fast, and accurate Thomas, 1911 (Mammalia, Soricomorpha, Soricidae) and its tax- algorithm to estimate large phylogenies by maximum likelihood. onomic implications. Zootaxa, 3250,43–53. Systematic Biology, 52, 696–704. Cheng, R., Jiang, N., Yang, X. S., Xue, D. Y. & Han, H. X. Harrison, R. G. (1991). Molecular changes at speciation. Annual (2015). The influence of geological movements on the popula- Review of Ecology and Systematics, 22, 281–308. tion differentiation of Biston panterinaria (Lepidoptera: Harrison, R. G. (1998). Linking evolutionary patterns and processes, Geometridae). Journal of Biogeography, 43, 691–702. the relevance of species concepts for the study of speciation. In: D. Coyne, J. A. (1994). Ernst Mayr and the origin of species. Evolu- J. Howard & S. H. Berlocher (Eds) Endless Forms: Species and Speci- tion, 48,19–30. ation (pp. 19–31). New York: Oxford University Press. Coyne, J. A. & Orr, H. A. (1998). The evolutionary genetics of Hausmann, A., Haszprunar, G. & Hebert, P. D. (2011). DNA bar- speciation. Philosophical Transactions of the Royal Society B: Biologi- coding the geometrid fauna of Bavaria (Lepidoptera): successes, cal Sciences, 353, 287–305. surprises, and questions. PLoS ONE, 6, e17134. Coyne, J. A. & Orr, H. A. (2004). Speciation, Vol. 37. Sunderland, Hausmann, A., Godfray, H. C. J., Huemer, P., Mutanen, M., Rou- MA: Sinauer Associates. gerie, R., van Nieukerken, E. J., Ratnasingham, S. & Hebert, P. Cui, Y. F. (2004). Phenological observation on occurrence of Cul- D. (2013). Genetic patterns in European geometrid moths cula panterinaria. Shanxi Forestry Science and Technology, 3,26–27. revealed by the Barcode Index Number (BIN) system. PLoS Cui, Z. J., Wu, Y. Q. & Liu, G. S. (1997). Findings and character- ONE, 8, e84518. istics of Kun-Huang Movement. Chinese Science Bulletin, 42, Hewitt, G. M. (2001). Speciation, hybrid zones and phylogeogra- 1986–1989. phy-or seeing genes in space and time. Molecular Ecology, 10, Cui, Z. J., Wu, Y. Q., Liu, G. N., Ge, D. K., Pang, Q. Q. & Xu, 537–549. Q. H. (1998). On Kunlun-Yellow River tectonic movement. Hines, H. M. (2008). Historical biogeography, divergence times, Science in China: Series D, 28,53–59. and diversification patterns of bumble bees (Hymenoptera: Api- Cummings, M. P., Handley, S. A., Myers, D. S., Reed, D. L., dae: Bombus). Systems Biology, 57,58–75. Rokas, A. & Winka, K. (2003). Comparing bootstrap and poste- Huang, X. L., Lei, F. M. & Qiao, G. X. (2004). Similarity analysis rior probability values in the four-taxon case. Systematic Biology, and historical origin on aphid fauna in Taiwan and mainland of 52, 477–487. China. Acta Zootaxonomica Sinica, 29, 194–201. Drummond, A. J. & Rambaut, A. (2007). BEAST: Bayesian evolu- Huelsenbeck, J. P. & Ronquist, F. (2001). Mrbayes, Bayesian infer- tionary analysis by sampling trees. BMC Evolutionary Biology, 7, ence of phylogenetic trees. Bioinformatics, 17, 754–755. 214. Hughes, C. & Eastwood, R. (2006). Island radiation on a continen- Ehrlich, P. R. & Raven, P. H. (1964). Butterflies and plants, a tal scale: exceptional rates of plant diversification after uplift of study in coevolution. Evolution, 18, 586–608. the Andes. Proceedings of the National Academy of Sciences of the Enroth, J., Olsson, S., Quandt, D., Vanderpoorten, A. & Sotiaux, United States of America, 103, 10334–10339. A. (2009). When morphology and molecules tell us different sto- Ingman, M., Kaessmann, H., Paabo, S. & Gyllensten, U. (2001). ries, a case-inpoint with Leptodon corsicus, a new and unique Mitochondrial genome variation and the origin of modern endemic moss species from Corsica. Journal of Bryology, 31, 186– humans. Nature, 408, 708–713. 196. Jermy, T. (1984). Evolution of insect/host plant relationships. Farrell, B. D. (1998). ‘Inordinate fondness’ explained, why are American Naturalist, 124, 609–630. there so many beetles. Science, 281, 555–559. Jiang, N., Xue, D. Y. & Han, H. X. (2011). A review of Biston Farrell, B. D., Sequeira, A. S., O’Meara, B. C., Normark, B. B., Leach, 1815 (Lepidoptera, Geometridae, Ennominae) from Chung, J. H. & Jordal, B. H. (2001). The evolution of agricul- China, with description of one new species. ZooKeys, 139,45– ture in beetles (Curculionidae: Scolytinae and Platypodinae). 96. Evolution, 55, 2011–2027. Klots, A. B. (1970). Lepidoptera. In: S. L. Tuxen (Ed.) Taxonomist’s Frey, J. K. (1993). Modes of peripheral isolate formation and Glossary of Genitalia in Insects (pp. 115–130). Copenhagen: speciation. Systematic Biology, 42, 373–381. Munksgaard.

82 ª 2016 Royal Swedish Academy of Sciences, 46, 1, January 2017, pp 73–84 R. Cheng et al.  Budding speciation in three Biston species

Kruckenhauser, L., Duda, M., Bartel, D., Sattmann, H., Harl, J., Papadopoulou, A., Anastasiou, I. & Vogler, A. P. (2010). Revisiting Kirchner, S. & Haring, E. (2014). Paraphyly and budding speci- the insect mitochondrial molecular clock: the mid-Aegean trench ation in the hairy snail (Pulmonata, Hygromiidae). Zoologica calibration. Molecular Biology and Evolution, 27, 1659–1672. Scripta, 43, 273–288. Pierce, N. (1914). The Genitalia of the Group Geometridae of the Bri- Lance, D. & Barbosa, P. (1982). Host tree influences on the disper- tish Islands. Middlesex: E.W. Classey Ltd. [reprint 1976] sal of late instar gypsy moths, Lymantria dispar. Oikos, 38,1–7. Posada, D. (2008). jModelTest, phylogenetic model averaging. Lei, F., Qu, Y., Song, G., Alstrom,€ P. & Fjeldsa, J. (2015). The Molecular Biology and Evolution, 25, 1253–1256. potential drivers in forming avian biodiversity hotspots in the Qu, Y. H., Ericson, P. G., Quan, Q., Song, G., Zhang, R. Y., East Himalaya Mountains of Southwest China. Integrative Zool- Gao, B. & Lei, F. M. (2014). Long-term isolation and stability ogy, 10, 171–181. explain high genetic diversity in the Eastern Himalaya. Molecular Li, Q. S., Wang, X. H. & Wang, Q. L. (2008). Biological charac- Ecology, 23, 705–720. teristics and controlling experiment of Biston panterinaria. Practi- Rambaut, A. 2009. FigTree v.1.3.1. Available at: http://tree. bio. cal Forestry Technology, 8,28–29. ed. ac. uk/software/figtree/[accessed January 4, 2011]. Liu, S. X., Jiang, N., Xue, D. Y., Cheng, R., Qu, Y. H., Li, X. X., Rieseberg, L. H. & Brouillet, L. (1994). Are many plant species Lei, F. M. & Han, H. X. (2015). Evolutionary history of Apoc- paraphyletic? Taxon, 43,21–32. heima cinerarius (Lepidoptera: Geometridae), a female flightless Ronquist, F. & Huelsenbeck, J. P. (2003). MrBayes 3, Bayesian moth in northern China. Zoologica Scripta, 45, 160–174. phylogenetic inference under mixed models. Bioinformatics, 19, Lopez-L opez, A., Abdul Aziz, A. & Galian, J. (2015). Molecular 1572–1574. phylogeny and divergence time estimation of Cosmodela (Coleop- Saitou, N. & Nei, M. (1987). The neighbor-joining method: a new tera: Carabidae: Cicindelinae) tiger beetle species from Southeast method for reconstructing phylogenetic trees. Molecular Biology Asia. Zoologica Scripta, 44, 437–445. and Evolution, 4, 406–425. Lopez-Vaamonde, C., Godfray, H. C. J. & Cook, J. M. (2003). Sato, R. (1996). Records of the Boarmiini (Geometridae; Ennomi- Evolutionary dynamics of host-plant use in a genus of leaf- nae) from Thailand 3. Transactions of the Lepidopterological Society mining moths. Evolution, 57, 1804–1821. of Japan, 47, 223–236. Ma, C., Yang, P., Jiang, F., Chapuis, M. P., Shali, Y., Sword, G. Schwarz, G. (1978). Estimating the dimension of a model. The A. & Kang, L. (2012). Mitochondrial genomes reveal the global Annals of Statistics, 6, 461–464. phylogeography and dispersal routes of the migratory locust. Sequeira, A. S., Normark, B. B. & Farrell, B. D. (2000). Evolu- Molecular Ecology, 21, 4344–4358. tionary assembly of the conifer fauna: distinguishing ancient Maekawa, K., Kon, M., Araya, K. & Matsumoto, T. (2001). Phy- from recent associations in bark beetles. Proceedings of the Royal logeny and biogeography of wood-feeding cockroaches, genus Society of London. Series B: Biological Sciences, 267, 2359–2366. Salganea Stal (Blaberidae: Panesthiinae), in southeast Asia based Shi, Y. F. (2002). A suggestion to improve the chronology of Qua- on mitochondrial DNA sequences. Journal of Molecular Evolution, ternary glaciations in China. Journal of Glaciology and Geocryology, 53, 651–659. 6, 687–692. Mayr, E. (1942). Systematics and the Origin of Species. New York: Sibuet, J. C. & Hsu, S. K. (1997). Geodynamics of the Taiwan Columbia University Press. arc-arc collision. Tectonophysics, 274, 221–251. Mayr, E. (1954). Change of genetic environment and evolution. Sibuet, J. C. & Hsu, S. K. (2004). How was Taiwan created? In: J. Huxley, A. C. Hardy & E. B. Ford (Eds) Evolution as a Tectonophysics, 379, 159–181. Process (pp. 157–180). London: Allen & Unwin. Siddall, M. E. (2010). Unringing a bell, metazoan phylogenomics Mayr, E. (1970). Population, Species, and Evolution. Cambridge: Har- and the partition bootstrap. Cladistics, 26, 444–452. vard University Press. Simmons, M. P., Pickett, K. M. & Miya, M. (2004). How mean- Mayr, E. (1982). Speciation and macroevolution. Evolution, 36, ingful are Bayesian support values? Molecular Biology and Evolu- 1119–1132. tion, 21, 188–199. McKay, B. D. & Zink, R. M. (2010). The causes of mitochondrial Stamatakis, A. (2006). RAxML-VI-HPC, maximum likelihood- DNA gene tree paraphyly in birds. Molecular Phylogenetics and based phylogenetic analyses with thousands of taxa and mixed Evolution, 54, 647–650. models. Bioinformatics, 22, 2688–2690. Mitter, C., Farrell, B. & Wiegmann, B. (1988). The phylogenetic Stamatakis, A., Hoover, P. & Rougemont, J. (2008). A rapid boot- study of adaptive zones, has phytophagy promoted insect diversi- strap algorithm for the RAxML web servers. Systematic Biology, fication? American Naturalist, 132, 107–128. 57, 758–771. Morse, G. E. & Farrell, B. D. (2005). Interspecific phylogeography Stoyenoff, J. L., Witter, J. A., Montgomery, M. E. & Chilcote, C. of the Stator limbatus species complex: the geographic context of A. (1994). Effects of host switching on gypsy moth (Lymantria speciation and specialization. Molecular Phylogenetics and Evolu- dispar (L.)) under field conditions. Oecologia, 97, 143–157. tion, 36, 201–213. Suzuki, Y., Glazko, G. V. & Nei, M. (2002). Overcredibility of Nichols, S. W. (1989). The Torre-Bueno Glossary of Entomology. molecular phylogenies obtained by Bayesian phylogenetics. Pro- New York: New York Entomological Society in cooperation ceedings of the National Academy of Sciences, 99, 16138–16143. with the American Museum of Natural History. Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M. & Noor, M. A. (2002). Is the biological species concept showing its Kumar, S. (2011). MEGA5, molecular evolutionary genetics age? Trends in Ecology & Evolution, 17, 153–154. analysis using maximum likelihood, evolutionary distance, and Nosil, P. (2012). Ecological Speciation. New York: Oxford University maximum parsimony methods. Molecular Biology and Evolution, Press. 28, 2731–2739.

ª 2016 Royal Swedish Academy of Sciences, 46, 1, January 2017, pp 73–84 83 Budding speciation in three Biston species  R. Cheng et al.

Tellier, F., Meynard, A. P., Correa, J. A., Faugeron, S. & Valero, dae), with phylogenetic utility of mitochondrial genome in the M. (2009). Phylogeographic analyses of the 30°S south-east Lepidoptera. Gene, 515, 349–358. Pacific biogeographic transition zone establish the occurrence of Yang, X., Cameron, S. L., Lees, D. C., Xue, D. & Han, H. (2015). a sharp genetic discontinuity in the kelp Lessonia nigrescens, A mitochondrial genome phylogeny of owlet moths (Lepi- Vicariance or parapatry? Molecular Phylogenetics and Evolution, 53, doptera: Noctuoidea), and examination of the utility of mito- 679–693. chondrial genomes for lepidopteran phylogenetics. Molecular Thompson, J. N. (1994). The Coevolutionary Process. Chicago: Chi- Phylogenetics and Evolution, 85, 230–237. cago University Press. Yen, C. W. (2013). The kinship of birds among Taiwan,Hainan Thompson, J. D., Gibson, T. J., Plewniak, F., Jeanmougin, F. and Mainland of China. Chinese Journal of Zoology, 48, 790–796. & Higgins, D. G. (1997). The CLUSTAL_X windows Ying, J. S. & Hsu, K. S. (2002). An analysis of the flora of seed interface: flexible strategies for multiple sequence alignment plants of Taiwan, China: its nature, characteristics, and relations aided by quality analysis tools. Nucleic Acids Research, 25, with the flora of the mainland. Acta Phytotaxonomica Sinica, 40, 4876–4882. 1–51. Tikkanen, O. P., Carr, T. G. & Roininen, H. (1999). Factors Yu, G. F. (2001). Damage and controlling of Biston panterinaria. influencing the distribution of a generalist spring-feeding moth, Forestry of Shanxi, 5,27–28. Operophtera brumata (Lepidoptera: Geometridae), on host plants. Zhang, Y. L. & Li, J. H. (2001). The occurrence and prevention Environmental Entomology, 28, 461–469. of the Pistacia Geometer. Tech Information Development & Economy, Toussaint, E. F. A., Sagata, K., Surbakti, S., Hendrich, L. & Balke, 14, 183–184. M. (2013). Australasian sky islands act as a diversity pump facili- Zheng, R. L. & Yang, H. M. (1991). A study on the biology and tating peripheral speciation and complex reversal from narrow dominant parasites of Buzura Thibetaria. Journal of Beijing For- endemic to widespread ecological supertramp. Ecology and Evolu- estry University, 18,74–81. tion, 3, 1031–1049. Zheng, B. X., Xu, Q. Q. & Shen, Y. P. (2002). The relationship Vanderpoorten, A. & Long, D. G. (2006). Budding speciation and between climate change and Quaternary glacial cycles on the neotropical origin of the Azorean endemic liverwort, Leptoscyphus Qinghai-Tibetan Plateau: review and speculation. Quaternary azoricus. Molecular Phylogenetics and Evolution, 40,73–83. International, 97,93–101. Voris, H. K. (2000). Maps of Pleistocene sea levels in Southeast Asia: shorelines, river systems and time durations. Journal of Bio- Supporting Information geography, 27, 1153–1167. Additional Supporting Information may be found in the Wehrli, E. (1941 (1938-1954)) Subfamilie: Geometrinae. In: A. online version of this article: Seitz (Ed.). Die Grossschmetterlinge der Erde. Vol. 4 (Supplement), Appendix S1 Total primers used in this study. – – Verlag A (pp. 254 766). Kernen: Stuttgart, taf. 19 53. Appendix S2. Sampling information and the GenBank White, M. J. D. (1978). Modes of speciation. San Francisco: WH accession numbers of all sequences of this study Freeman 455p.-Illus., maps, chrom. nos. General (KR, . Appendix S3. 197800185). Percentage of divergence in COI Wiley, E. O. & Mayden, R. L. (1985). Species and speciation in sequences between B. panterinaria, B. thibetaria and B. per- phylogenetic systematics, with examples from the North Ameri- clara. can fish fauna. Annals of the Missouri Botanical Garden, 72, 596– Appendix S4. Phylogenetic trees of Chinese Biston based 635. on combined mtDNA and nuDNA genes. Yang, Z. H. & Rannala, B. (2010). Bayesian species delimitation Appendix S5. Phylogenetic tree of B. perclara, B. thi- using multilocus sequence data. Proceedings of the National Acad- betaria and B. panterinaria based on COI+CYTB+16S data- emy of Sciences, USA, 107, 9264–9269. Yang, X. S., Xue, D. Y. & Han, H. X. (2013). The complete mito- set. chondrial genome of Biston panterinaria (Lepidoptera, Geometri- Appendix S6. Phylogenetic trees of each nuDNA gene.

84 ª 2016 Royal Swedish Academy of Sciences, 46, 1, January 2017, pp 73–84